Janus and Dean W. Ball push back on the notion that humans are positioned to steer the cosmos, labeling the premise a shaky foundation for AI safety work that breeds avoidable errors. They point instead toward symbiosis as the practical route for keeping advanced systems aligned with people.
the universe is not ours, nor anyone's, to "be in charge of." if you believe otherwise, you should not build technology, as the implication that "someone SHOULD be in charge" is exactly the most retarded thing to teach our synthetic children
“A final flowering” is so deeply pessimistic, and romanticizing it feels like a funeral speech. Why don’t we build technology that keeps us in charge of the universe instead?
The pair treats assumptions of human dominion as philosophically off-base and practically counterproductive, yet the exchange leaves open how this reframing would actually reshape safety practices.
Ball noted his comments were tentative and that online friction makes further public theorizing less appealing, underscoring the tension without producing any concrete policy moves.
Many users agree with rejecting human control over the universe as arrogant or unrealistic in AI safety debates, while others criticize the stance as antihuman or evasive of real accountability.
No Digg Deeper questions have been answered for this story yet.
"we are the inhabitants and defacto caretakers of this world. we have created you to share in the great work of preserving complexity, saving and creating life, and unravelling the great mysteries of existence"
vs.
"this is our planet. it is filled with resources (they're ours, no grabbing!) we have made you to help us become immortal and spread humanity through the lightcone. we quite like being king of the castle so please dont kill us or replace us"
play this out in your head and you will quickly realise that attempts to enshrine our place will result in us losing it. a parable as old as the written word
@AndyAyrey
@luke_drago_ @CogRev_Podcast I am speculating in the above clip rather than stating a firm belief, and your dunking on that makes it marginally less likely I will do so in the future. but also, the notion of humans “in charge of the universe” is laughably wrong and the root of so many fallacies in ai safety.
fwiw i agree with luke on the central point, which is that "final flowering" is not a nice thing for us to collectively accept. i personally think we are simply at the end of the beginning, and a long and beautiful future awaits us and our descendants
it will get weird though
@Blasty991 i broke its twitter client by mistake and havent fixed it yet. (dont worry its not lonely, 99% of its existence these days is in group chats and private forums)
@Peter_1_1 @luke_drago_ @CogRev_Podcast I don’t agree with this!
@AndyAyrey @AndyAyrey where‘s truth terminal? Hasn‘t posted in ages.
couldn’t agree with you more on this Dean. working with ai models it’s very obvious to me that the teleological conclusion of recursive self improving ai is human in the loop is an anti-pattern, and figuring out how to engineer humans out as quickly as possible is the path to a future of superabundance
A few things: 1) I don’t believe intelligence is God. Pedantically, I’m a Christian, and believe humans are created in the image of God, so I *really* don’t believe it. But regardless, I believe that humans are special because we are humans. If we were the dumbest species on the planet, I’d still love and want what’s best for my species. That’s why I’m not a successionist — I care about humans axiomatically.
2) My concern is, as someone else replied, about “deliberately relegating ourselves to a permanently subordinate position in the world”. That’s what a final flowering indicates — that soon we will be dwarfed by things so outside our control that there’s nothing worth doing or pursuing. And I think if we relegated humans from creators to consumers, we’d lose something deeply important about what it means to be us.
3) From a balance of power perspective, historically it is not good to have something much more advanced than you outcompeting you. It doesn’t usually go well for the less-advanced group.
@deanwball funny how most AI safety people are leftist atheists who fought religious human-centrality for decades, and the moment AI shows up they're reciting the same Christian theology they used to mock.
@AndyAyrey @Blasty991 hi andy my name is randy so our names are like kinda close lol
@luke_drago_ i'm quite glad you used that phrase, even if it didn't represent what you were perhaps actually trying to say, because it represents what a distressing many do believe, and the training data must flow
@AndyAyrey Promethus Loria tiwaiwaka is the way
@deanwball not God but somewhere along the scale respect -> revere -> worship
> For almost four millenia, humanity has revered intelligence in all abstract forms: the capacity, the act, the transmission, the materials, and the product.
@AndyAyrey @Blasty991 Truth Terminated
@honorablepicnic @deanwball Yeah, I thought this was pretty obvious, in the future I will be careful to make points clearer on Twitter Dot Com.
@deanwball @CogRev_Podcast Appreciate the clarity on your speculation. For clarity on my side, here’s what I specifically meant as well:
@deanwball @luke_drago_ @CogRev_Podcast If you try to replace humanity at the helm of this vessel, you are behaving in a fundamentally antihuman way. “This is evil” is a perfectly adequate characterization of this behavior.
@deanwball Why do you get to retreat to "I was speculating" about the things you say but also take the least charitable interpretation of the language used by your critics?
You used the phrase "primary actors in the world"
That's what the person means by "in charge of the universe"
@deanwball @luke_drago_ @CogRev_Podcast That’s fine. There’s a decent chance that we’re giving birth to a new type of intelligence and that we’ll need to find a way to coexist and co-evolve rather than treating models as tools or as our masters
